[canvas] Improve (Closes #7145)

youtube_dl/extractor/canvas.py

@@ -1,80 +1,65 @@
-# coding: utf-8
 from __future__ import unicode_literals
 
-import os
-import urlparse
-
-from youtube_dl import utils
 from .common import InfoExtractor
+from ..utils import float_or_none
 
 
 class CanvasIE(InfoExtractor):
-    _VALID_URL = r'(?:https?://)?(?:www\.)?canvas\.be/video/(?P<id>.+)'
+    _VALID_URL = r'https?://(?:www\.)?canvas\.be/video/(?:[^/]+/)*(?P<id>[^/?#&]+)'
     _TEST = {
         'url': 'http://www.canvas.be/video/de-afspraak/najaar-2015/de-afspraak-veilt-voor-de-warmste-week',
         'md5': 'ea838375a547ac787d4064d8c7860a6c',
         'info_dict': {
-            'id': 'de-afspraak/najaar-2015/de-afspraak-veilt-voor-de-warmste-week',
-            'title': 'De afspraak veilt voor de Warmste Week',
+            'id': 'mz-ast-5e5f90b6-2d72-4c40-82c2-e134f884e93e',
+            'display_id': 'de-afspraak-veilt-voor-de-warmste-week',
             'ext': 'mp4',
-            'duration': 49,
+            'title': 'De afspraak veilt voor de Warmste Week',
+            'description': 'md5:24cb860c320dc2be7358e0e5aa317ba6',
+            'thumbnail': 're:^https?://.*\.jpg$',
+            'duration': 49.02,
         }
     }
 
     def _real_extract(self, url):
-        video_id = self._match_id(url)
+        display_id = self._match_id(url)
 
-        webpage = self._download_webpage(url, video_id)
+        webpage = self._download_webpage(url, display_id)
 
         title = self._search_regex(
-            r'<h1 class="video__body__header__title">(.+?)</h1>', webpage,
-            'title')
-        data_video = self._html_search_regex(
-            r'data-video=(["\'])(?P<id>.+?)\1', webpage, 'data-video', group='id')
-        json_url = 'https://mediazone.vrt.be/api/v1/canvas/assets/' + data_video
-        data = self._download_json(json_url, video_id)
+            r'<h1[^>]+class="video__body__header__title"[^>]*>(.+?)</h1>',
+            webpage, 'title', default=None) or self._og_search_title(webpage)
+
+        video_id = self._html_search_regex(
+            r'data-video=(["\'])(?P<id>.+?)\1', webpage, 'video id', group='id')
+
+        data = self._download_json(
+            'https://mediazone.vrt.be/api/v1/canvas/assets/%s' % video_id, display_id)
 
         formats = []
         for target in data['targetUrls']:
-            if 'type' and 'url' in target:
-                extension = utils.determine_ext(target['url'])
-                if target['type'] == 'PROGRESSIVE_DOWNLOAD':
-                    formats.append({
-                        'format_id': extension,
-                        'url': target['url'],
-                        'protocol': 'http',
-                    })
-                elif target['type'] == 'HLS':
+            format_url, format_type = target.get('url'), target.get('type')
+            if not format_url or not format_type:
+                continue
+            if format_type == 'HLS':
                 formats.extend(self._extract_m3u8_formats(
-                        target['url'], video_id, entry_protocol='m3u8_native',
-                        ext='mp4',
-                        preference=0,
-                        fatal=False,
-                        m3u8_id='hls'))
-                elif target['type'] == 'HDS':
+                    format_url, display_id, entry_protocol='m3u8_native',
+                    ext='mp4', preference=0, fatal=False, m3u8_id=format_type))
+            elif format_type == 'HDS':
+                formats.extend(self._extract_f4m_formats(
+                    format_url, display_id, f4m_id=format_type, fatal=False))
+            else:
                 formats.append({
-                        'format_id': extension,
-                        'url': target['url'],
-                        'protocol': 'HDS',
+                    'format_id': format_type,
+                    'url': format_url,
                 })
-                elif target['type'] == 'RTMP':
-                    formats.append({
-                        'format_id': extension,
-                        'url': target['url'],
-                        'protocol': 'rtmp',
-                    })
-                elif target['type'] == 'RTSP':
-                    formats.append({
-                        'format_id': extension,
-                        'url': target['url'],
-                        'protocol': 'rtsp',
-                    })
-
         self._sort_formats(formats)
-        duration = utils.int_or_none(data.get('duration')) / 1000
+
         return {
             'id': video_id,
+            'display_id': display_id,
             'title': title,
+            'description': self._og_search_description(webpage),
             'formats': formats,
-            'duration': duration,
+            'duration': float_or_none(data.get('duration'), 1000),
+            'thumbnail': data.get('posterImageUrl'),
         }